Forecasting 1

Journal of Business & Economic Studies, Vol. 17, No.
2, Fall 2011
47

Intangible Investments and the Pricing of Corporate SGA Expenses

Rongbing Huang, Kennesaw State University
Gim S. Seow, University of Connecticut
Joe Z. Shangguan, Robert Morris University

Abstract

This study examined whether the market fully prices the reported Selling, General, and
Administrative (SGA) expenses when this item includes an intangible investment component.
For a sample of intangible investment-intensive firms, we showed that their SGA expenses
benefit future operating performances. Evidence suggests some degree of market inefficiency in
the pricing of SGA expenses and the intangible investment component. Furthermore, the
financial analysts do not appear to appreciate fully the future benefits of the component in their
earnings forecasts. Finally, the pertinent disclosures in firms annual reports are so inadequate as
to attenuate the market mispricing, suggesting a significant room for future improvement.

Keywords: Selling, General, and Administrative Expenses, Intangible Investment, Mispricing,
Analyst Forecast, Disclosure
JEL Classification: M41, G10, G14

Introduction

Accounting standard setting often entails balancing between two primary information
qualities, relevance and reliability
1
, to accommodate the needs of different firm stakeholders.
This balanced consideration may sometimes lead to the expensing, rather than capitalizing, of
certain value-relevant firm expenditures. For instance, despite the Financial Accounting
Standards Boards (FASB) acknowledgement that expenditures for R&D constitute a significant
element of the United States economy and are vital for its growth, Financial Accounting
Standard (FAS) No. 2 mandates that all research and development (R&D) costs be charged to
expense as incurred on the ground of uncertainty about the future benefits of individual R&D
projects and the difficulty involved in assigning a causal relationship between expenditures and
benefits. The same rationale underlies the full expensing, rather than either full or partial
capitalizing, of some other firm expenditures that are important to firms long-term success,
thereby producing relatively lower reported profits and net assets.
The purpose of this paper was to examine empirically the market pricing of a summary
expense item on the income statement
2
, Selling, General, and Administrative (SGA) expenses,
on the premise that it may conceal some capital investment-like expenditures for developing
intangibles. The concealment raises the possibility that market participants fixate on the SGA
expenses per se and fail to recognize fully the long-term value implications of the intangible
investment component. A similar phenomenon was documented by Sloan (1996) in which
investors tended to fixate on reported earnings while failing to recognize the differential value
implications of its two components: cash and accruals. Prior research has documented that stock
Journal of Business & Economic Studies, Vol. 17, No. 2, Fall 2011
48

prices reflect the value of R&D capital and that market valuation is meaningfully related to the
fundamental benefits and risk of R&D (Chan, Sougiannis, & Lakonishok, 2001; Lev &
Sougiannis, 1996; Lev & Zarowin, 1999). However, research also provided evidence of either a
systematic mispricing of R&D or a compensation for an extra risk factor associated with R&D
(Lev & Sougiannis) and evidence of higher return volatility for R&D-intensive firms (Chan et
al.), suggesting an informational cost in the equity market associated with the R&D accounting
treatment. We extended this line of research and examined whether similar informational cost
exists for other investment-like components of the SGA expenses.
Many firms economic rents stem from such intangibles as knowledge, intellectual
capital, organizational capital, and customer loyalty. Most of these value-relevant intangibles,
however, do not meet the asset recognition criteria and often are deemed as practically infeasible
to be recognized as assets (Upton, 2001). As a result, the various firm expenditures aimed at or
effectively generating these intangibles in house often are accounted for as operating expenses
3

and incorporated into the SGA expenses on the income statement. Given the dollar-by-dollar
downward effect of SGA expenses on earnings and the importance of intangibles to firms value
creation, this study investigated whether the stock market can see through the conservative
accounting treatment and overcome the general lack of pertinent firm disclosures and
appropriately price the SGA expenses reported on the income statement.
To gain insight into this question, it would have been ideal to be able to measure
accurately the intangible investment component of SGA expenses for each firm and to examine
directly the cross-sectional association between stock prices (returns) and SGA expenses, with or
without the intangible investment component included. However, this approach was implausible
because firm-level data on intangible investment outlays were unavailable. Unlike the R&D
expenditures, which are required to be disclosed, firms rarely disclose some other expenditures
that are expensed pursuant to accounting standards but may effectively generate intangibles.
Consequently, a realistic situation would be that investors perceive the inclusion of the intangible
investment component in some firms SGA expenses accounts, but do not know its exact
amount. Our approach, accordingly, was to examine a group of firms that were characterized by
this kind of valuation situation
4
.
The sample we examined was based on the lists of 500 best information technology (IT)
firm users published annually by the InformationWeek magazine. These firms not only invested
heavily on information systems, but also made substantial complementary spending in areas such
as employee training, work process redesign, and organizational reshuffling (Brynjolfsson &
Hitt, 2000). The majority of these expenditures are summed into the SGA expenses for financial
reporting purpose. In effect, however, they should be regarded as investment in intangibles
insofar as they create a crucial source of firm valueorganizational capital (Brynjolfsson & Hitt,
2000; Lev & Radhakrishnan 2004; Prescott & Visscher, 1980).
We first documented that the SGA expenses of the Informationweek500 firms, because
they included a significant intangible investment component, had a positive impact on future
operating performance, in contrast with a larger industry and size-matched sample. We found a
set of consistent evidence that suggested some degree of market mispricing of SGA expenses for
the InformationWeek500 firms. The stock market seemed to underestimate the contribution of
the IT and other complementary spending to equity value. We also examined whether financial
analysts fully impute the implications of the intangible investment component of SGA expenses
for future profitability. The results suggested that these experts either fail to do so or appear to be
49

conservative in their earnings forecasts. Furthermore, the current level of firm disclosure in the
annual reports was so inadequate as to attenuate the market mispricing.
Our study is of potential interest to both firms and investors. We documented the market
mispricing of a major information item on the income statement, the selling, general and
administrative expenses, for the InformationWeek500 firms. This group of firms invests heavily
internally to develop intangibles that are increasingly important in todays economy. With the
arguable limitations to the accounting treatment for this type of investment, firms need to find
alternative ways to reduce the informational cost related to investors difficulty in understanding
the value relevance of the intangible investment. This need seems particularly imperative, given
our evidence that the current level of firm disclosure is inadequate. For investors, our study
provides some practical guidance as to what type of firms are most likely to have an intangible
investment component concealed in the SGA expenses account.

Sample and Data

Our sample firms came from those identified by InformationWeek magazine as the most
innovative and effective corporate IT users. The magazine conducts yearly surveys of both
public and private U.S. firms with relatively intensive IT spending. Firms are evaluated and
assigned scores in such areas as IT budgets, technology deployment, E-business, customer
knowledge, infrastructure, and business and technology strategies. Around each September, the
magazine publishes a list of top 500 firms based on the overall scores. We chose these firms to
construct our sample because their reported SGA expenses were most likely to incorporate a
significant intangible investment component.
To gauge the extent of intangible investment that may be concealed in the SGA expenses,
we examined the limited information available on IT spending by the InformationWeek500 firms
provided by the magazine. Table 1 shows that the average IT budgets were $442 million in 2000
and $334 million in 2004, representing 4.31% and 3.68% of revenues, respectively. These
budgets, if all treated as expenses
5
, would have accounted for about 19.82% and 18.58% of SGA
expenses (excluding R&D expenditures), respectively. Moreover, besides IT spending, firms
often made complementary investments in areas such as employee training, work process
redesign, and organization restructuring. These intangible investments that were accounted for as
SGA expenses typically exceeded the IT spending. Brynjolfsson et al. (2002) documented that
the widespread use of information technology had increased investments in intangible
organizational assets. They used firm-level data and found that each dollar of installed computer
capital in a firm was associated with at least five dollars of market value, after controlling for
other assets. They interpreted this value as evidence of a large stock of intangible assets that
complemented the computer investment.
Note that the IT budgets shown in Table 1 were only summary survey data; the exact
amounts of IT spending, as well as other complementary intangible investments by individual
firms, usually are unavailable to investors, because most firms do not report such information.
Investors often have to rely on sporadic disclosures from various sources to infer the extent of IT
and other complementary investments. For example, Owens & Minor Inc., an
InformationWeek500 firm, provided the following disclosure in its 2002 annual report:
To support its strategic efforts, the company has developed information systems to
manage virtually all aspects of its operations, including warehouse and inventory
management, asset management and electronic commerce. In July 2002, the company
50

entered into a new, seven-year information technology agreement with Perot Systems
Corporation, expanding an existing outsourcing relationship.

Table 1. InformationWeek500 Firms Dollars Spent on IT
2000 2001 2002 2003 2004
Average company revenue
(Billion) $10.26 $12.47 $9.43 $9.65 $9.09
Average dollars spent on IT
(Million) $442 $484 $320 $353 $334
Average IT budget as a % of
revenue 4.31% 3.88% 3.39% 3.66% 3.68%
Average IT budget as a % of
SGA expenses (excluding R&D
expenditures)* 19.82% 23.17% 18.85% 18.58% 18.58%
Source: InformationWeek magazine
* Computed based on sample means of SGA expenses

Even this kind of information disclosure is rare among the firms
6
. This example represents the
current information environment in which investors value the reported SGA expenses and the
intangible investment component.
To form our sample, we started with the 4,500 InformationWeek500 firms during 1996
through 2004. From them, we obtained 1,187 firm-year observations after eliminating firms not
on Compustat and CRSP and without data on total assets, sales, SGA expenses, and operating
income. We used this sample to examine the implications of SGA expenses for future operating
performance. The sample size varied after imposing additional data requirements in the market
efficiency tests and analysts forecasts efficiency test.
Panel A in Table 2 presents basic accounting and market information for the
InformationWeek500 sample firms. These firms tended to be large, with median sales of $4.48
billion and market capitalization of $4.73 billion. The median SGA expenses to sales ratio was
0.16 and both median return on assets and median profit margin were 0.10. Panel B in Table 2
shows that the sample distribution across years was largely even. As for industry distribution
(untabulated), the firms spanned a broad range of 52 two-digit SIC industries, with the most
observations from Chemicals and Pharmaceuticals (SIC code 28, N=117 or 9.8%) followed by
Machinery and Computer equipment (SIC code 35, N=107 or 9.0%). Therefore, our results were
unlikely to be biased towards any particular years or industries.

51

Table 2. Sample Descriptive Statistics
Panel A: Accounting and Market Profiles of InformationWeek500 Sample Firm-Years
Variable N Mean Median Std Dev Min. Max.
Total assets ($mil) 1,187 15,408 4,566 39,211 243 479,921
Sales ($ mil.) 1,187 10,576 4,481 20,426 157 192,319
Common equity ($
mil.) 1,187 4,001 1,659 6,794 3 78,927
Market value ($ mil.) 1,187 16,610 4,726 39,126 20 460,304
SGA expenses / Sales 1,187 0.18 0.16 0.11 0.00 0.70
Return on assets 1,187 0.11 0.10 0.08 (0.34) 0.57
Profit margin 1,187 0.11 0.10 0.12 (1.02) 0.53
Market model beta 1,187 1.00 0.94 0.49 0.01 4.00
Number of employees
(thousand) 1,174 40.65 19.70 61.56 1.18 608.00

Panel B: Distribution of Sample Firms Across Years
Year N %
1996 115 9.69
1997 117 9.86
1998 112 9.44
1999 109 9.18
2000 147 12.38
2001 152 12.81
2002 150 12.64
2003 144 12.13
2004 141 11.88
Total 1,187 100.00

Empirical Analyses

The Implications of SGA Expenses for Future Operating Performance
Our analysis started with the examination of the impact of SGA expenses on future
operating performance. Prior research suggested that the time-series behavior of earnings was by
and large a random walk or a random walk with drift (Ball & Watts, 1972; Freeman, Ohlson, &.
Penman, 1982), namely, the next-period earnings could best be predicted by the current-period
earnings. Mozes (1992) suggested that the random walk earnings model could be expanded to an
AR (2) model, which also includes the last-period earnings:
t t t t
u Earnings a Earnings a a Earnings + + + =
+ 1 2 1 0 1
. To examine whether the SGA expenses had any
additional predictive value for future earnings, we added it into the earnings forecast model and
scaled all the variables by total assets (sales). This modification brought us to estimating the
following two equations:

52

t t t t t
u SGA a ROA a ROA a a ROA + + + + =
+ 3 1 2 1 0 1
(3.1a)

t t t t t
e SGA b PM b PM b b PM + + + + =
+ 3 1 2 1 0 1
, (3.1b)

where ROA is return on assets measured as operating income (OI) (Compustat annual data item #
178) divided by average total assets (Data #6), PM is profit margin measured as operating
income divided by sales (Data #12), and t indicates fiscal year t. SGA is SGA expenses
excluding R&D expenditures (Data #189 Data #46) and is divided by average total assets in
(3.1a) and by sales in (3.1b).
If SGA indeed included only expenses, then by the nature of expense, it should not have
affected future performance and we would have expected a
3
or b
3
to be insignificant. However, if
SGA also included a portion of expenditures that could effectively generate intangible assets and,
hence, impact a firms long-term performance, as was likely the case for the
InformationWeek500 firms, then we would have expected a
3
and b
3
to be positive.
The estimation results shown in Table 3 confirmed our expectations. In Panel A, we first
estimated (3.1a) and (3.1b) for the InformationWeek500 sample. Both a
3
(estimate=0.025, t-
stat.=3.30) and b
3
(estimate=0.111, t-stat.=3.48) were positive and significant at the 1% level,
suggesting that for the InformationWeek500 firms, the SGA expenses account had a component
that benefits future performance.
As a comparison, we estimated the same equations for a sample of firms from Compustat
that were matched with the InformationWeek500 firms by industry and firm size during the same
period (1996-2004)
7
. The results were distinctively different. In the return on assets regression,
a
3
was positive but insignificant. In the profit margin regression, b
3
was even significantly
negative. The results suggested that for the larger population of firms, overall, SGA expenses do
not benefit future performance, consistent with the accounting definition of expenses.
Panel B of Table 3 provides the estimation results for the pooled sample. We introduced a
dummy variable, DUM
t
, for firms in the InformationWeek500 sample and an interaction variable,
SGA
t
DUM
t
. The models were as follows:

t t t t t t t t
u DUM a DUM SGA a SGA a ROA a ROA a a ROA + + + + + + =
+ 5 4 3 1 2 1 0 1
(3.1c)

t t t t t t t t
e DUM b DUM SGA b SGA b PM b PM b b PM + + + + + + =
+ 5 4 3 1 2 1 0 1
(3.1d)
In the above equations, coefficients
3
a or
3
b alone represents the effect of SGA expenses
on future performance for the non-InformationWeek500 firms, while ) (
4 3
a a + or ) (
4 3
b b +
measures the effect for the InformationWeek500 firms. The results in Panel B. confirmed those in
Panel A. In the return on assets regression,
3
a was positive but insignificant,
4
a was positive and
significant at the 5% level. In the profit margin regression,
3
b was significantly negative, but
4
b was much more positive and significant at the 1% level. Once again, the results suggested that
the SGA expenses had a positive impact on future performance due to an intangible investment
component for the InformationWeek500 firms but not for the larger population of firms.

53

Table 3 The Implication of SGA Expenses for Future Operating Performance
Panel A: Separate OLS Estimations for the InformationWeek500 Sample and the
Matched Sample Based on Industry and Size
t t t t t
u SGA a ROA a ROA a a ROA + + + + =
+ 3 1 2 1 0 1

InformationWeek500 sample Matched sample
Parameter Estimate Adjusted t-stat. Estimate
Adjusted t-
stat.
a
0
0.012 (3.85)*** 0.017 (9.56)***
a
1
0.893 (15.47)*** 0.619 (11.64)***
a
2
-0.082 (-1.49) 0.025 (1.29)
a
3
0.025 (3.30)*** 0.020 (1.28)
N 1,187 13,066
Adjusted R
2
72.2% 50.4%

t t t t t
e SGA b PM b PM b b PM + + + + =
+ 3 1 2 1 0 1

InformationWeek500 sample Matched sample
Parameter Estimate Adjusted t-stat. Estimate
Adjusted t-
stat.
b
0
0.002 (0.37) 0.041 (4.18)***
b
1
0.837 (9.16)*** 0.684 (10.07)***
b
2
-0.038 (-0.35) -0.088 (-3.38)***
b
3
0.111 (3.48)*** -0.072 (-2.42)**
N 1,187 13,742
Adjusted R
2
72.5% 46.4%

(Table continues)

54

Table 3 (continued)

Panel B Pooled OLS estimations
(a)
t t t t t t t t
u DUM a DUM SGA a SGA a ROA a ROA a a ROA + + + + + + =
+ 5 4 3 1 2 1 0 1

(b)
t t t t t t t t
e DUM b DUM SGA b SGA b PM b PM b b PM + + + + + + =
+ 5 4 3 1 2 1 0 1

(a) (b)
Parameter Estimate
Adjusted t-
stat. Parameter Estimate
Adjusted t-
stat.
a
0
0.016 (9.38)*** b
0
0.040 (4.17)***
a
1
0.627 (12.02)*** b
1
0.688 (10.28)***
a
2
0.023 (1.18) b
2
-0.088 (-3.41)***
a
3
0.020 (1.28) b
3
-0.070 (-2.37)**
a
4
0.031 (2.53)** b
4
0.219 (5.20)***
a
5
0.006 (1.56) b
5
-0.023 (-3.17)***
N 14,253 14,929
Adjusted R
2
51.4% 46.9%
** and *** indicate statistical significance at the 5% and 1% level, respectively. White
(1980) t-statistics are used.
Variable definition (for corresponding fiscal years):
ROA - return on assets measured as operating income divided by average total assets
PM - profit margin percentage measured as operating income divided by sales
SGA - selling, general and administrative expenses (excluding R&D) scaled by average
total assets in the ROA regression and by sales in the PM regression
DUM - a dummy variable equal to 1 for the InformationWeek500 observations and 0
otherwise

Table 4 provides further evidence of the positive impact of SGA expenses on future
performance for the InformationWeek500 sample. We laid out a common measure of operational
productivity, sales per employee, for the year (year 0) when a firm was selected into the
InformationWeek500 list and the four subsequent years. As the table shows, overall sales per
employee increased over time for both mean (from $305,310 in year 0 to $341,010 in year 4) and
median (from $214,070 in year 0 to $237,770 in year 4). In each of the five years, the
InformationWeek500 firms had higher sales per employee than the median firms in their two-
digit industries. For industry-adjusted sales per employee, the median increased (from $22,520 in
year 0 to $29,630 in year 4) while the mean decreased (from $93,440 in year 0 to $84,340 in year
4) over time. These results also suggested heavy IT spending by the InformationWeek500 firms
because IT is often a substitute for human labor. The difference appeared quite distinctive.

55

Table 4. Sales per Employee (in $1,000) in and After the Year When Firms
Are Selected Into the InformationWeek500 List. (Year 0 Is When a Firm Is
Selected Onto the InformationWeek500 List (N=646))
Sales per employee Mean Median
Year 0 305.31 214.07
Year 1 317.84 220.87
Year 2 319.48 224.59
Year 3 319.05 232.19
Year 4 341.01 237.77
Industry-adjusted sales per employee
Year 0 93.44 22.52
Year 1 97.58 24.63
Year 2 92.48 24.30
Year 3 82.19 28.83
Year 4 84.34 29.63
Variable definition:
Sales per employee - sales divided by number of employees
Industry-adjusted sales per employee - firm sales per employee minus two-digit
industry median sales per employee

The Market Pricing of SGA Expenses for InformationWeek500 Firms
In this section, we built on the above evidence that for the InformationWeek500 firms the
SGA expenses had a positive effect on future operating performance due to an intangible
investment component (IT and complementary spending) and examined whether the stock
market fully prices this effect.
We conducted this analysis in two steps. First, we examined whether the stock market
perceives the reported SGA expenses as carrying value-relevant information. A standard
procedure in the finance literature is to examine whether there is any association between
contemporaneous stock returns and the variable of interest (i.e., the SGA expenses) while
controlling for factors known to affect returns (Fama & French, 1993; Rajgopal, Shevlin, &
Venkatachalam, 2003). Accordingly, we estimated the following equation:

t t t t t t t
SGA EP BTM MV BETA SAR c | | | | | o + + + + + + =
5 4 3 2 1
ln ln (3.2a)

where SAR is size-adjusted abnormal return measured as the firm's raw buy-and-hold return for
the 12-month period ending 3 months after fiscal year-end minus the buy-and-hold return on a
size-matched portfolio during the same period, BETA, lnMV, and lnBTM, represented the three
Fama-French factors: systematic risk, size, and book-to-market equity ratio. BETA was estimated
from the market model using 60 monthly returns prior to year t (at least 24 monthly returns were
required). lnMV was measured as the natural logarithm of market value at the fiscal year-end.
lnBTM was measured as the natural logarithm of book-to-market equity ratio. EP was the
earnings-price ratio (year-end stock price at t-1 is used) intended to control for potential E/P ratio
anomaly (Basu, 1977). SGA was the variable of interest. A significant coefficient on SGA
56

implied that it contains value-relevant information reflected in the contemporaneous returns.
Thus, we expected
5
| to be significantly positive.
A complementary test of market efficiency was to regress subsequent returns (
1 + t
SAR ) on
the same set of explanatory variables. Specifically, we estimated the following equation:

t t t t t t t
SGA EP BTM MV BETA SAR c | | | | | o + + + + + + =
+ 5 4 3 2 1 1
ln ln (3.2b)

A significant coefficient (
5
| ) on SGA would have implied that the market was correcting
itself by incorporating value-relevant information contained in SGA in the subsequent period;
hence, the absence of full efficiency in the concurrent period.

We ran three alternative estimations to ensure the robustness of results. Panel A of Table
5 reports the simple OLS estimation results. The estimates for
5
| are -0.277 in the
contemporaneous returns regression and 0.226 in the subsequent returns regression, both were
significant at the 10% level. Panel B reports heteroskedasticity-consistent results using nonlinear
OLS estimation, while Panel C reports results based on the Newey-West (1987) estimation
method to adjust for potential heteroskedasticity and serial correlation in error terms. Results in
Panels B and C are largely consistent with those in Panel A. We considered our results as
evidence of market inefficiency. The market does not appear to price fully the positive impact of
SGA expenses on future operating performance for the InformationWeek firms, as documented
in Table 3.

57

Table 5. Regressions of Contemporaneous and Subsequent Stock Returns
on SGA Expenses (N=941)
Model:
t t t t t t t t
e SGA EP BTM MV BETA SAR SAR + + + + + + =
+ 5 4 3 2 1 1
ln ln or | | | | | o
Panel A: Simple OLS estimation
SAR
t
SAR
t+1

Parameter Estimate T-stat. Estimate T-stat.
0.146 1.33 -0.039 -0.43
1
0.099 2.68
***
0.035 1.14
2
-0.027 -2.18
**
-0.011 -1.06
3
-0.136 -5.45
***
0.004 0.20
4
-0.703 -1.80
*
1.440 4.42
***
5
-0.277 -1.71
*
0.226 1.68
*
Adj. R
2
(%) 4.1 2.7

Panel B: Heteroskedasticity-Consistent Nonlinear OLS Estimation
SAR
t
SAR
t+1

Parameter Estimate Adjusted t-stat. Estimate Adjusted t-stat.
0.146 1.44 -0.039 -0.42
1
0.099 2.08
**
0.035 1.03
2
-0.027 -2.12
**
-0.011 -1.06
3
-0.136 -2.88
***
0.004 0.21
4
-0.703 -1.95
*
1.440 3.57
***
5
-0.277 -1.51 0.226 1.67
*
Adj. R
2
(%) 5.0 3.1
Panel C: Nonlinear GMM Estimation Using Newey-West Method
SAR
t
SAR
t+1

Parameter Estimate Adjusted t-stat. Estimate Adjusted t-stat.
0.140 1.71
*
-0.039 -0.44
1
0.091 2.56
**
0.035 1.46
2
-0.025 -2.49
**
-0.011 -1.15
3
-0.129 -2.79
***
0.004 0.28
4
-0.716 -1.96
*
1.440 3.45
***
5
-0.265 -1.74
*
0.226 2.43
**
Adj. R
2
(%) 4.6 2.6
*, ** and *** indicate statistical significance at the 10%, 5% and 1 % level, respectively.
Variable definition:
SAR - size-adjusted abnormal return measured as the firm's raw buy-and-hold return for the 12-
month period ending 3 months after fiscal year-end (of t or t+1) minus the buy-and-hold return on
a size-matched portfolio during the same holding period
BETA - beta estimated from the market model using 60 monthly returns prior to year t (at least 24
monthly returns are required)
lnMV - natural logarithm of market value
lnBTM - natural logarithm of book-to-market equity ratio
EP - earnings-to-price ratio (year-end stock price at t-1 is used)
SGA - selling, general, administrative expenses (excluding R&D) scaled by sales

58

Mishkin Test of Market Efficiency in Pricing SGA Expenses
In this sub-section, we add further evidence on whether the market fully prices the
intangible investment component of the SGA expenses by using an alternative method. Recently,
researchers have adopted the Mishkin (Abel & Mishkin, 1983; Mishkin 1983,) method to test
market efficiency in a variety of settings. For example, Sloan (1996) used it to test whether the
market is efficient in pricing the differential implications of the accrual and cash components of
earnings for future earnings. Rajgopal et al. (2003) used this method to test whether the market
fully appreciates the implications of firms backlogs for future earnings. Similar to prior studies,
we specified the models as follows:

t t t t
v SGA OIB OI + + + =
+ 2 1 0 1
(3.3a)
t t t t t
u SGA OIB OI SAR + =
+ +
) (
*
2
*
1 0 1 1
| (3.3b)

where OI is operating income scaled by average total assets at the beginning and the end of the
fiscal year, OIB is operating income before SGA expenses scaled by average total assets, namely,
OIB = (OI + SGA). SAR and SGA are as defined before. Equation (3.3a) is referred to as the
forecasting equation. It estimates the forecasting ability (or persistence) of the two components
of earnings for future earnings, as represented by
1
and
2
. The purpose of decomposing OI into
OIB and SGA was to facilitate the examination of market pricing of the implications of SGA
expenses for future earnings. This was in the same spirit as Sloan (1996), in which earnings were
purposefully decomposed into accrual and cash components in order to examine their differential
persistence. Equation (3.3b) could be referred to as the rational pricing equation. According to
this model, the market reacts to unexpected change in earnings, namely, OI
t+1
-expectation of
OI
t+1
, where the expectation of OI
t+1
is simply
t t
SGA OIB
2 1 0
+ + based on equation (3.3a).
Market efficiency imposes that
*
1 1
= and
*
2 2
= . In other words, it imposes a market
rationality constraint by allowing investors to anticipate the implications of the two earnings
components for future earnings. A rejection of the joint constraints would indicate market
inefficiency.
The two equations, (3.3a) and (3.3b), were estimated jointly using iterative weighted non-
linear least squares method (Abel & Mishkin, 1983; Mishkin 1983). The restrictions imposed by
market efficiency were tested using a likelihood ratio statistic that is distributed
asymptotically ) (
2
q _ : ) / ln( 2
u r
SSR SSR n , where q is the number of restrictions imposed by
market efficiency, n is the number of observations, SSR
r
is the sum of squared residuals from the
restricted weighted system, and SSR
u
is the sum of squared residuals from the unrestricted
weighted system.

59

Table 6. Mishkin Test of Market Efficiency in the Pricing Implications of SGA Expenses
for Subsequent Earnings (N=1,125)
Nonlinear generalized least squares regression is applied to the following equations
system:
(1)
t t t t
v SGA OIB OI + + + =
+ 2 1 0 1

(2)
t t t t t
u SGA OIB OI SAR + =
+ +
) (
*
2
*
1 0 1 1
|

Parameter Estimate
Asymptotic
Standard Error

t-stat.
0
0.008 0.002 3.66
***
1
0.807 0.017 49.06
***
1
* 0.915 0.075 12.28
***
2
-0.773 0.020 -37.96
***
2
* -0.983 0.106 -9.30
***
2.259 0.343 6.59
***
*** indicates statistical significant at the 1% level.
Test of market efficiency:
*
1 1
= and
*
2 2
=
Likelihood ratio statistic: 9.25
Marginal significance level: 0.01
OI- operating income (Compustat data item # 178) scaled by average total assets
SGA - selling, general, and administrative expenses scaled by average total assets
OIB - OI before SGA expenses scaled by average total assets, namely, OI+SGA
SAR - size-adjusted abnormal return measured as the firm's raw buy-and-hold return for
the 12-month period ending 3 months after fiscal year-end (of t+1) minus the buy-and-
hold return on a size-matched portfolio during the same holding period

Table 6 reports the estimation results. The joint hypothesis
*
1 1
= and
*
2 2
= was
rejected at the 1% significance level (likelihood ratio statistic = 9.25). The estimate of
1
was
0.807, while the estimate of
*
1
is 0.915. This implied that the stock market perceived the OIB
component of earnings as being more persistent than suggested by the time-series forecasting
equation (i.e., 3.3a). Turning to SGA, its coefficient estimate from equation (3.3a)
2
was equal
to -0.773, while from equation (3.3b)
*
2
was equal to -0.983, indicating that the market
perceived the SGA component of earnings as being more persistent than suggested by equation
(3.3b). Meanwhile,
*
2
= -0.983 meant that the market treats SGA expenses as reducing firm
value nearly dollar by dollar, This generally may be true but was inconsistent with the evidence
in Section 3.1 that SGA expenses of the InformationWeek500 firms encompassed significant
capital investments. Therefore, the Mishkin test provided further evidence that the market did not
fully price the value implications of the SGA expenses for the InformationWeek500 firms.

Does Firm Disclosure Help Attenuate the Market Inefficiency?
The mispricing of SGA expenses documented above could be attributable to investors
difficulty in interpreting the reported amounts. The difficulty arises from both the full expensing
accounting rule and the current status of inadequate firm disclosure regarding the intangible
60

investment component of SGA expenses. As we pointed out at the beginning, the FASB may
have valid reasons to mandate full expensing of expenditures whose outcomes were highly
uncertain. The investors thus relied more on firms providing adequate information disclosure.
We then examined whether the current disclosures in firms annual reports improved market
valuation.
We based our analysis on equations (3.2a) and (3.2b) and introduced a new variable for
firm disclosure on IT investment and other related spending. The model specification was as
follows:

t t t t
t t t t t t t
e it discl it discl SGA
SGA EP BTM MV BETA SAR SAR
+ + +
+ + + + + =
+
_ _ *
ln ln or
7 6
5 4 3 2 1 1
| |
| | | | | o
, (3.4)

where it discl _ is a dummy variable for disclosure of IT-related information in management
discussions of SGA expenses and all other variables are as defined previously. The coding of
it discl _ required reading each sample firm's annual report. We randomly selected 10% (93)
firms from the sample used for equation (3.2b) that had accessible annual reports in the EDGAR
database. We then read these firms annual reports with primary focus on the discussions of SGA
expenses in the Management Discussion and Analysis (MD&A) section and the footnotes to the
financial statements. If there was any discussion of IT and other complementary spending or any
description of IT programs, it discl _ was coded as 1, otherwise 0 was assigned. In total, we
identified 15 firms with IT disclosures in the discussions of SGA expenses. For example,
Magnetek, Inc. had the following disclosure in its 1997 annual report:
Selling, general and administrative (SG&A) expense was $159.9 million (13.4% of net
sales) in fiscal 1997 compared to $164.9 million (14.2% of net sales) in fiscal 1996
While the Company continues to review opportunities to reduce support costs, expenses
associated with upgrades in information systems, quality programs and organizational
capability will limit the ability to reduce SG&A expense in fiscal 1998.
The interaction variable in equation (3.4), it discl SGA _ * , is the variable of interest.
Theory (Verrecchia, 2001) and empirical evidence have suggested that voluntary disclosure
helps reduce information asymmetry. If investors are informed that SGA expenses encompass IT
and complementary spending, they would weigh SGA expenses less negatively when forecasting
future earnings. Thus, ceteris paribus, the market inefficiency documented in the preceding
sections will be attenuated for the disclosing firms. For this reason, we expect the coefficient on
it discl SGA _ * to be negative (i.e., 0
6
< | ).

61

Table 7. The Effect of Disclosure About IT Investment on the Market Pricing
Adjustment Relating to SGA Expenses (N=93)
Model:
t t t t
t t t t t t
e it discl it discl SGA SGA
EP BTM MV BETA SAR or SAR
+ + + +
+ + + + =
+
_ _
ln ln
7 6 5
4 3 2 1 1
| | |
| | | | o

SAR
t
SAR
t+1

Parameter Estimate t-stat. Estimate t-stat.
0.209 0.69 -0.072 -0.25
1
0.084 0.81 0.024 0.25
2
-0.034 -1.08 -0.012 -0.40
3
-0.200 -2.51
**
-0.036 -0.48
4
-0.865 -0.78 1.870 1.79
*
5
-0.812 -1.72
*
-0.129 -0.29
6
1.038 0.91 0.775 0.72
7
-0.181 -0.84 -0.170 -0.83

Adj. R
2
(%) 3.8 -2.4
* and **indicate statistical significance at the 10 % and 5% level, respectively
discl - a dummy variable equal to 1 if a firm provides voluntary disclosure on IT investment in
connection with SGA expenses in its annual report of year t, and 0 otherwise
Other variables are as defined in Table 5.

Table 7 reports the OLS estimation results for the combined sample (disclosure and
nondisclosure firms). In the contemporaneous returns ( )
t
SAR regression, the coefficient on SGA,
5
| , was negative and significant at the 10% level. Inconsistent with our expectation, the
coefficient on it discl SGA _ * ,
6
| , was positive but insignificant. In the subsequent returns
( )
1 + t
SAR regression, neither
5
| nor
6
| was significant. We concluded that the IT-related
disclosures in firms' discussions of SGA expenses in annual reports did not provide much useful
information to investors; thus, they did not help mitigate market mispricing. It is likely that firms
may have disclosed IT-related information through other channels. Nevertheless, our
examination of annual reports revealed that firms disclosures were often too brief and
qualitative in nature, leaving significant room for future improvement on this information
channel.

Do Analysts Fully Appreciate the Investment Nature of the Spending Reported as SGA
Expenses?
We examine whether financial analysts fully appreciate the capital investment nature of
spending reported as SGA expenses. Financial analysts act as information intermediaries in the
stock market. Compared to average investors, they have developed strong financial expertise
through training and experience and have greater access to corporate information. Insights into
how analysts incorporate information contained in SGA expenses can help us better understand
the market mispricing of SGA expenses.
62

To see how analysts fare, we adopted a method similar to that in Rajgopal et al. (2003),
who examined how analysts weigh information contained in firms order backlogs in their
earnings forecasts. Specifically, we estimated the following three regressions:

t t t t
e SGA EPS EPS + + + =
+ 2 1 0 1
| | | (3.5a)
t t t t
u SGA EPS FEPS + + + =
+ 2 1 0 1
(3.5b)
) ( ) ( ) ( ) (
2 2 1 1 0 0 1 t t t t t
e u SGA EPS FERROR + + + =
+
| | | (3.5c)

where EPS is reported earnings per share, excluding extraordinary items obtained from
Compustat (item #58), FEPS is I/B/E/S consensus forecast of earnings per share reported in the
fourth month after fiscal year-end. The lag of three to four months ensures that SGA expenses
information is available to analysts. Both EPS and FEPS are scaled by the stock price on the last
trading day of fiscal year t. FERROR is equal to FEPS - EPS and stands for analysts forecast
error. SGA is as defined before.
Equation (3.5a) was a time-series forecast model for reported earnings. Coefficients
1
|
and
2
| captured the information content of current earnings and SGA expenses in the time-
series forecast of future earnings. Coefficients
1
and
2
in equation (3.5b), on the other hand,
represented the weights analysts assigned to the two variables in their forecasts of future
earnings. Whether analysts properly assign the weights was seen by comparing the two pairs of
coefficients. In particular, if
2 2
| < (both coefficients should be negative because SGA
represents expenses), it meant that analysts overestimated the negative implications of the
current reported SGA expenses for future earnings. On the contrary, analysts underestimated it.
Equation (3.5c) was obtained by subtracting equation (3.5a) from equation (3.5b). This
procedure allowed the statistical testing of the coefficient differences (
1
-
1
| ) and (
2
-
2
| ). We
expected (
2
-
2
| ) < 0. Although analysts have expertise, the lack of adequate disclosure on the
intangible investment component may still have impaired their ability to make effective
assessment about the implications of SGA expenses for future earnings.
Table 8 reports the estimation results for (3.5a) - (3.5c) based on a sample of 625 firms
with necessary data available from Compustat and I/B/E/S. Examining the coefficient on
t
EPS ,
Panel A shows
1
| =0.340 (t-stat.=11.94). Panel B shows
1
=0.199 (t-stat.=4.72). The difference
(
1
| -
1
) was statistically significant as shown in Panel C (t-stat.=-3.00), indicating that analysts
underestimated the persistence of earnings. This result was consistent with prior research
(Ahmed et al. 2002, Rajgopal et al. 2003). More interestingly, the coefficient on SGA
2
| =-0.021
(t-stat.=-1.21) in Panel A and
2
=-0.080 (t-stat.=-3.10) in Panel B, and Panel C shows the
difference (
2
-
2
| ) =-0.058 was statistically significant at the 5% level (t-stat.=-2.02). These
results, consistent with our expectation, indicated that analysts assigned a more negative weight
on SGA expenses in their forecasts than that implied by the time-series forecast model. Like
average investors, analysts failed to fully appreciate the benefits of IT investment and other
complementary spending concealed in the reported SGA expenses
8
. This result was consistent
with the finding of Lev and Radhakrishnan (2004) that financial analysts earnings forecasts did
not fully reflect the value of organizational capital resulted from IT and other related
investments. Given this inability of analysts, it was not surprising that the overall market
misprices the SGA expenses reported by the InformationWeek500 firms.
63

Table 8. Analysis of Whether Analysts Fully Appreciate the Information
Contained in SGA Expenses in Their Earnings Forecasts (N =625)
The following equations were estimated:
(1)
t t t t
e SGA EPS EPS + + + =
+ 2 1 0 1
| | |
(2)
t t t t
e SGA EPS FEPS + + + =
+ 2 1 0 1
| | |
(3) ) ( ) ( ) ( ) (
2 2 1 1 0 0 1 t t t t t
e u SGA EPS FERROR + + + =
+
| | |

Parameter Estimate t-stat
Panel A:
Equation
(1)
Adj. R
2
=
19.4%
0
0.025 6.52
***

1
0.340 11.94
***

2
-0.021 -1.21

Panel B:
Equation
(2) Adj. R
2
= 5.4%
0
0.059 10.5
***

1
0.199 4.72
***

2
-0.080 -3.10
***

Panel C:
Equation
(3) Adj. R
2
= 1.5%
0 -
0
0.034 5.42
***

1 -
1
-0.142 -3.00
***

2-
2
-0.058 -2.02
**

**, and *** indicate statistical significance at the 5% and 1 % level, respectively
EPS- earnings per share excluding extraordinary items (Compustat item #58) scaled
by the previous year-end share price
SGA - selling, general, and administrative expenses (excluding R&D) scaled by sales
FEPS - I/B/E/S consensus forecast of earnings per share reported in the fourth month
after fiscal year end scaled by the previous year-end share price
FERROR - analysts' forecast error, equal to FEPS - EPS

Conclusion

The SGA expenses reported on many firms income statements included substantial
expenditures with a capital investment nature because they effectively may have generated
intangibles that are crucial for firms long-term success. The full expensing of these expenditures
and the lack of pertinent disclosure could create a difficulty for investors valuation, thereby
causing an informational cost in the equity market.
64

Our evidence was consistent with this view. For a sample of InformationWeek500 firms
that were perceived as intensive in IT and other complementary investments, we found that these
firms SGA expenses were positively associated with future operating performance, a
phenomenon that did not exist for the larger population of other firms. However, the market did
not seem to fully price the value implication of the intangible investment component of the SGA
expenses or the item as a whole. Neither did the financial analysts fully appreciate the
implication in their earnings forecasts. Our results also showed that the current level of
disclosure in firms annual reports was not enough to attenuate this particular market mispricing.
Two caveats are worth mentioning regarding the interpretation of our results. First, our
results should not be over-generalized. We do not infer that for all firms the reported SGA
expenses encompassed an intangible investment component or that there was a market
mispricing of their value implications. Our analyses apply best to those firms that rely more on
intangibles to compete and need to invest heavily to develop them. Second, we do not view our
results as a favor for proposing a change in the current accounting standard, namely, the full
expensing of such intangible investment-like expenditures, although it contributes to an
informational cost in the equity market. As mentioned in the beginning, accounting standard
setting often requires a delicate balancing between different stakeholders. Our results, however,
do suggest that there is a significant room for improving the pertinent disclosures on the firms
part.

End Notes

1
Financial Accounting Standards Board (FASB) Concepts Statement No. 2
2
Another major summary expense item on the income statement is Cost of Sales.
3
We do not deny that a portion of these expenditures may qualify as capital investment.
4
In another approach, Kovacs (2004) and Shangguan (2005) extract a capital investment component out
of SGA expenses and find that this component is positively associated with future operating performance.
They first estimate the industry-level amortization rates of SGA expenses by regressing earnings on a
chain of current and lagged SGA expenses using large cross-sectional samples. The capital investment
component is measured as the sum of the unamortized SGA expenses. This estimated capital component
is then used to test the market valuation of firms' investment-type spending. However, their approach is
not without limitations. First, a large cross-sectional sample may include many firms for which the SGA
expenses account contains little or no capital investment components. Second, the amortization rates of
SGA expenses estimated based on large sample cross-sectional regression are invariant at least for firms
within an industry. Thus, their approach does not help distinguish different intensities of spending on
intangibles for firms within an industry.
5
It is without a doubt that firms would capitalize a portion of IT spending such as purchase of computers.
Our point, as we will illustrate immediately, is that at least another, and often significant, portion of IT
spending, along with other investment-like expenditures, are accounted for as operating expenses.
6
This view is partly substantiated by our investigation of firm disclosures discussed in Section 3.4. Only
15 out of 93 sample firms briefly mention the IT component in their discussion of SGA expenses in
annual reports. No firm discloses specific amounts of the IT and complementary spending.
7
We first sort all available Compustat firms sales into five quintiles. The firm-years of those non-
InformationWeek500 firms with the same two-digit SIC codes and sales quintiles as those of the
InformationWeek500 firms are used to form the matched sample.
8
Another potential explanation is that analysts are conservative in their forecasts and valuation due to the
high uncertainty associated with IT investment outcome. Dehning et al. (2006), for example, show that
IT spending increases earnings forecast dispersion and error, which in turn translates into lower market
value for the firm.
65

References

Abel, A., & Mishkin, F. (1983, January). An integrated view of tests of rationality, market
efficiency, and the short run neutrality of monetary policy. Journal of Monetary
Economics, 11, 3-24.
Ball, R., & Watts, R. (1972). Some time series properties of accounting income. Journal of
Finance, 27(3), 663-681.
Basu, S. 1977. The investment performance of common stocks in relation to their price to
earnings ratios: A test of the efficient markets hypothesis. Journal of Finance, 32, 663-
662.
Brynjolfsson, E., & Hitt, L. M. (2000). Beyond computation: Information technology,
organizational transformation and business performance. Journal of Economic
Perspectives, 14(4), 23-48.
Brynjolfsson, E., Hitt, L. M., & Yang, S. (2002). Intangible assets: Computers and organization
capital. Brooking Papers on Economic Activity, 1, 137-198.
Chan, L., Sougiannis, T., & Lakonishok, J. (2001). The stock market valuation of research and
development expenditures. Journal of Finance, 56(6), 2431-2456.
Fama, E., & French, K. (1993). Common risk factors in the returns on stocks and bonds, Journal of
Financial Economics, 33, 3-56.
Financial Accounting Standards Board. (1974). Financial Accounting Standard No. 2:
Accounting for research and development costs. Norwalk, CT: FASB.
Financial Accounting Standards Board. (1985). Financial Accounting Standard No. 86:
Accounting for the costs of software to be sold, leased, or otherwise marketed. Norwalk,
CT: FASB.
Freeman, R., Ohlson, J., & Penman, S. (1982, Autumn). Book rate-of-return and prediction of
earnings changes: An empirical investigation. Journal of Accounting Research, 20, 3-42.
Kovacs, G. J. (2004). The future benefits in selling, general and administrative expenses.
Doctoral dissertation, Columbia University.
Lev, B., & Radhakrishnan, S. (2004). The measurement of firm-specific organization capital.
NBER Working Paper No. 9581.
Lev, B., & Sougiannis, T. (1996). The capitalization, amortization, and value-relevance of R&D.
Journal of Accounting and Economics, 21, 107-138.
Lev, B., & Zarowin, P. (1999). The boundaries of financial reporting and how to extend them.
Journal of Accounting Research, 37(2), 353-385.
Mishkin, F. (1983). A rational expectations approach to macroeconometrics: Testing policy
effectiveness and efficient markets models. Chicago, IL: University of Chicago Press for
the National Bureau of Economics Research.
Mozes, H. (1992). Modeling growth in the annual earnings time series. Journal of Business
Finance and Accounting, 19, 817-837.
Newey, W. K. & West, K. D. (1987). A simple, positive semi-definite, heteroskedasticity and
autocorrelation consistent covariance matrix. Econometrica, 55(3), 703-708.
Prescott, E. C., & Visscher, M. (1980). Organizational capital. Journal of Political Economy, 88,
446461.
Rajgopal, S., Shevlin, T., & Venkatachalam, M. (2003). Does the stock market fully appreciate
the implications of leading indicators for future earnings? Evidence from order backlog.
Review of Accounting Studies, 8, 461-492.
66

Shangguan, Z. (2005). Intangible investments and the implied cost of equity capital. Doctoral
dissertation, University of Connecticut.
Sloan, R. G. (1996). Does stock price fully reflect information in accruals and cash flows about
future earnings? The Accounting Review, 71(3), 289-315.
Upton, Jr., W. S. (2001). Business and financial reporting, challenges from the new economy.
Special report. Financial Accounting Standards Board, USA.
Verrecchia, R. (2001). Essays on disclosure. Journal of Accounting and Economics, 32, 97180.

97

About the Authors

Rongbing Huang, Ph.D., is an Associate Professor of Finance at Kennesaw State University.
He obtained Ph.D. in Finance from the University of Florida. His research focuses on corporate
finance and investments, especially investment banking and capital structure. He has published
in the Journal of Financial and Quantitative Analysis (JFQA), Journal of Corporate Finance,
and Financial Markets, Institutions & Instruments. He received the JFQA William F. Sharpe
Award in 2009.

Gurumurthy Kalyanaram (GK), Ph.D., received his Ph.D. from MIT Sloan School of
Management. He is a management consultant and currently the Dean of Amrita School of
Business. He has published in a variety of journals including International Journal of Research
in Marketing, Journal of Marketing Research, Journal of Consumer Research, Marketing
Science, and Review of Industrial Organization. For his services, GK has been recognized by
MIT with the Harold Lobdell Award. For his research contributions, GK also has been
recognized by American Marketing Association and INFORMS. He serves on the editorial
boards of several scholarly journals.

Doh-Khul Kim., Ph.D., is Associate Professor of Economics and Finance at North Central
College; prior to that, he was Associate Professor at Mississippi State University. He received his
Ph.D. from the University of Georgia. He has published in the journal International Advances in
Economic Research.

Zakia Mishal, Ph.D., is Professor of Economics at the Department of Economics, faculty of
Business and Economics at Yarmouk University, Jordan. She holds her MA and Ph.D. degrees in
Economic Development and Urban Economics from the University of Illinois at Chicago, USA.
Her research and teaching interests include sources of economic development, international
development, urbanization, and the role of women in the economy.

John J. Phelan, Ph.D., received his B.S. in business and economics, 65 and MA in economics,
67, from Indiana University and Ph.D. from The George Washington University, 77; senior
economist, Federal Trade Commission, 1969-81; Associate Executive Secretary to the Secretary,
U.S. Dept of Health and Human Services; 1981-89 (regulatory policy advisor). While working
for the Secretary of HHS, Dr. Phelan was involved in FDA policy issues and in particular, the
one that resulted in the elimination of prohibitions against the advertising of Rx drugs directly to
consumers. Dr. Phelan currently teaches economics at the University of New Haven with
research interests in the costs and benefits of Rx advertising, especially with respect to life-
saving drugs such as the statin (anti-cholesterol) drug class.

Stuart Rosenberg, Ph.D., is an Associate Professor in the Leon Hess Business School of
Monmouth University in West Long Branch, New Jersey. He formerly was on the faculty of
Dowling College from 2000-2010. He held management positions in the financial services
industry for over twenty years, first at Manufacturers Hanover Trust Company and later at First
Chicago Corporation. He has written both empirical papers and management case studies that
have been published in various academic journals. Dr. Rosenberg earned a B.A. from Marquette
98

University; an M.A. from the University of Wisconsin-Madison; an A.P.C. from New York
University; and an M.B.A. and Ph.D. from Fordham University.

Gim S. Seow, Ph.D., is an Associate Professor of Accounting at the University of Connecticut.
His research interests include accounting standard-setting and securities regulation in
international markets, accounting for financial derivatives, corporate risk management policies,
audit quality and industry expertise, valuation of intangible assets, and so forth. He has published
in several journals, including the Journal of Accounting and Economics, Contemporary
Accounting Research, and Accounting, Organizations and Society.

Joe Z. Shangguan, Ph.D., is an Associate Professor of Accounting at Robert Morris University.
He received his Ph.D. in Accounting from the University of Connecticut. His recent research
interests include intangible assets valuation, goodwill impairment, and mergers and acquisitions.
He has published in journals such as the Journal of Corporate Finance and Review of
Quantitative Finance and Accounting.

Jeungbo Shim, Ph.D., is Assistant Professor of Business Administration at Illinois Wesleyan
University. He received his Ph.D. from Georgia State University. His primary research interests
are mergers and acquisitions, capital structure, diversification, and risk measures. His research
has been published in Journal of Banking and Finance, Journal of Financial Services Research,
and International Journal of Business and Economics.

Michaeline Skiba, Ph.D., is an Associate Professor in the Leon Hess Business School of
Monmouth University in West Long Branch, New Jersey. In the business sector, she held senior
management positions within three Fortune 500 companies, where she developed marketing and
management programs, pharmaceutical marketing strategy (pre-launch), healthcare symposia and
colloquia (for CME and CPE credit), journal supplements, and market research (focus groups
and telephone-based interviews). Dr. Skiba earned a B.S. in education and biological sciences
from Loyola University Chicago; an M.S.I.R. from Loyola University Chicago; an M.Ed. from
Boston College; and an Ed.D. from Columbia University.

Reproducedwith permission of thecopyright owner. Further reproductionprohibited without permission.

Forecasting 1

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Forecasting 1

Hochgeladen von

Copyright:

Verfügbare Formate

Journal of Business & Economic Studies, Vol. 17, No.

Das könnte Ihnen auch gefallen